[Wikipedia-l] Vote on voting method for final round

[Tomasz Wegrzanowski]

On Fri, Sep 05, 2003 at 07:10:04AM -0700, Jimmy Wales wrote:
> Tomasz Wegrzanowski wrote:
> > Condorcet hasn't caused any trouble for Debian, and it assured fair
> > elections without any tactical voting and less prefered outcomes being
> > selected due to peculiarities of voting process.
> 
> I, too, am a fan of Condorcet, despite the complexity.  HOWEVER, for
> the PRESENT election, I would support any sensible and easy method,
> because it's just a logo, and any of the top competitors are going to
> be excellent.
> 
> Condorcet is sufficiently complex to make voting by hand and
> tabulating by hand really difficult.  But I would support a simple
> implementation of Condorcet in the software, as I think that Condorcet
> really is superior to other methods precisely in helping to find a
> _consensus_, i.e. the avoidance of divisive tactical voting.

There are many simple programs to calculate Condorcet winner, like this one:
http://www.eskimo.com/~robla/politics/condorcet-program.html
Debian has Condorcet software integrated with GPG-signed voting by mail,
but it's a bit too heavy procedure for us.

The "harder" part is a parser for ballots, finding the winner is trivial -
just filling NxN table and making a summary of it.